[RFC PATCH 4/4] linux: remove time64 redirects for __TIMESIZE=64 architectures
Arnd Bergmann
arnd@arndb.de
Wed Nov 11 23:22:07 GMT 2026
On Wed, Nov 11, 2026, at 17:46, Adhemerval Zanella Netto wrote:
> On 11/11/26 12:49, Arnd Bergmann wrote:
>> On Wed, Nov 11, 2026, at 14:07, Arnd Bergmann wrote:
>
> The --enable-kernel just define which __ASSUME_* macros are set in
> kernel-features.h and the ENOSYS fallbacks. It also sets the NT_GNU_ABI_TAG
> note, although I think it is just used by ld.so and file to prints the expected
> kernel version.
>
> And the minimal requires kernel has been raised before (7aab07e4e57,
> 6dad2c06889, d0f5b3f851c, 5b4ecd3f956, 139ace95756).
Ah, so historically a six year old kernel regularly fell out of
support, but then it remained at v3.2.
> Rising could be interesting to remove some fallbacks, for instance:
>
> * Up to 4.4/4.11: the fallbacks for renameat2 (3.15), execveat (3.19),
> mlock2 (4.4) and statx (4.11). Also the arch socketcall gates: 4.3 on i386,
> s390 and m68k, and 4.4 on sparc32. Each is one or two files.
>
> * 5.1 (the largest cleanup): __ASSUME_TIME64_SYSCALLS becomes unconditional
> on 32-bit ABIs. It also drops the ipc() multiplexer path on i386, s390,
> m68k and similar arches.
Right, at that time we tried to synchronize all syscall numbers and
also added a couple of missing ones along with the time64 and ipc calls:
- io_uring, pidfd_send_signal, pkey and rseq everywhere
- get{egid,euid,ppid}, umount2, shmat, io_pgetevents on alpha
- statx on sh
- seccomp on m68k
> * 5.3 to 6.6: clone3 (5.3), which removes the clone3_supported probe in
> clone-internal.c:76. Then faccessat2 (5.8), close_range (5.9),
> FUTEX_LOCK_PI2 (5.14) and fchmodat2 (6.6).
clone3 is still missing on sh and hexagon and was only recently added
on sparc (and nios2). We have assigned the number since it's common for
all but the actual implementation required work by the architecture
maintainers.
> It does come with some costs though:
>
> * Containers on old hosts, and this is the main reason the floor is still low.
> RHEL 7 is 3.10, RHEL 8 is 4.18 (maintained until 2029), Ubuntu 20.04 is 5.4,
> and Amazon Linux 2 is 4.14 or 5.10. Any floor above 4.18 drops RHEL 8 hosts.
As far as I can tell, these are all only for 64-bit architectures (x86, arm64,
powerpc, s390), right? Raising the minimum kernel version for the 32-bit
architectures would be enough for dropping the !__ASSUME_TIME64_SYSCALLS
fallback handling while keeping native containers on older enterprise distros
working as before.
What this would not cover is future i386 containers on those old x86_64
hosts. I don't know if that is still relevant, I see that there are a lot
of i386 images on Docker Hub, but those are all based on Alpine/musl
rather than glibc.
Arnd
More information about the Libc-alpha
mailing list